Glossary
AI interviewer
An AI interviewer is software that runs an interview on an employer’s behalf: it asks the questions, keeps time and, in many tools, asks follow-up questions and writes up the results with scores.
“AI interviewer” covers a range of tools. At one end are chatbots that ask a few typed questions and tick boxes. At the other are voice tools that hold a spoken conversation, ask more when an answer is vague, and write up what they heard, skill by skill.
What an AI interviewer does
- Asks your questions, either word for word or as a starting point for a conversation.
- Keeps time, so a 15-minute interview stays 15 minutes.
- Asks follow-up questions when an answer is thin, if the tool supports it.
- Records the conversation and turns it into written text.
- Rates the answers against the skills you choose, often with a score for each skill.
What it shouldn’t do
It shouldn’t make the hiring decision. It also shouldn’t judge things that have nothing to do with the job, like facial expressions, tone of voice, the room behind someone, or how “confident” they look. These say little about how well someone will do the job, and they make it easy to build unfair treatment into the tool. A sound setup follows a simple rule: the AI prepares, a person decides.
An example
A startup is hiring five sales reps whose job is to find and contact possible customers. It sets up an interview with four skills: researching possible customers, responding when a customer pushes back, writing follow-up messages, and bouncing back from rejection. The AI interviewer talks with each of 120 applicants over a week. The founder reads the top 25 write-ups, checks the quoted evidence, watches a few recordings and invites 10 people to a call.
Questions to ask the company selling the tool
- What exactly is scored, and what isn’t?
- Can I see the reasoning and evidence behind every score?
- Can I change any result, and is that change recorded?
- How are applicants told they’re talking to an AI?
- Does language, accent or grammar affect scores?
- Is applicant data used to train AI models?
- What happens when the sound quality is poor?
How it differs from AI note-takers
Some AI tools sit in on interviews run by people and take notes. They help the human interviewer. An AI interviewer takes the interviewer’s place for one interview, usually the first, so the team’s time goes to later ones. See how AI interview scoring works for more on how answers are rated.